Interactive visual preview of data helps users to quickly identify data of interest in large and complex data sets. Visual preview of visual data, such as images, video and graphic files, has been widely used in many applications, including the Windows operating system from Microsoft Corp. For data that does not have an established visual representation (i.e., non-visual data), however, such as many forms of tabular and database data, visual preview of data is seldom used. Visualizing non-visual data requires users with visualization expertise to choose an appropriate interactive visual form. In addition, visualizing non-visual data requires an implementation that can ensure the necessary interactive speed for large data sets.
There is an increasingly large class of users, however, for whom automatic visual preview of non-visual data would be valuable. These users, often referred to as “everyday business users,” typically have extensive domain knowledge but are typically not visualization or computer experts. Yet as part of their daily responsibilities, they perform situational analysis tasks over large amounts of non-visual data (such as sales data in a database). For such data, visual preview can be of great benefit for them to quickly browse data and find interesting content. It is difficult for everyday business users to create visual preview of data that is tailored to a user's context. A need therefore exists for automatic visual preview of non-visual data.